The University of Texas at Austin, in collaboration with consultants from the Pacific Institute on Research and Evaluation, is submitting this proposal to obtain support for an exploratory study to merge Texas driving record data and Texas treatment data to identify and evaluate the risk and protective factors for DWI/DUI recidivism among individuals who have ever received alcohol and/or drug treatment. The study has two aims: (1) To examine the strength of factors predicting DWI recidivism among individuals who have ever received substance abuse treatment, and (2) To create profiles of individual in treatment that predict risk of DWI recidivism based on the interaction of demographics, substance abuse factors, and driving factors. In addition to studying substance-related and driving-related factors, we will investigate recidivism rates for deterrents such as DWI education classes and interlocks for all Texas drivers, as well as rates for those who have been in treatment. Several factors are hypothesized to interact to predict the risk of recidivism: demographics, underlying severity of substance disorder, previous risky driving behaviors, intensity of substance treatment, and legal deterrents (e.g. interlocks, DWI education, etc.). We propose to disentangle these factors to create a predictive profile of who is most likely to recidivate and determine the interventions that work best for people with different combinations of characteristics or backgrounds. We will merge well over 8 years of longitudinal data on driving records (over 94,000 cases per year, which contain DWI arrests and penalties, accident reports, and other driving citations issued) with treatment data (over 80,000 admissions per year) on primary, secondary, and tertiary substances of abuse, frequency of use, history of treatment episodes, and severity and mental health problems as well as demographic information and status at discharge and follow-up, including abstinence.